This was originally posted on November 27, 2010 with different photos and no comparison with other animals. Since I now have far more photographs, I have decided to re-post some of the old horse color genetics posts with better photos.
The base colors of horses are bay, black and chestnut, possibly with the addition of wild bay and seal brown (tan-point.) These colors are distinguished by where red and black pigment are found, both where on the whole horse and where on individual hairs. I’m going to go into more detail this time on what determines these base colors.
Red pigment in horses (more correctly, phaeomelanin) can appear brownish red to copper, sometimes approaching gold, in the absence of dilution factors. With dilution factors, it can include white, cream, tan, yellow and gold shades.
Black pigment (more correctly, eumelanin) is black in the absence of modifying genes. In horses, the genes that dilute black to blue-gray or black to chocolate brown are not known to occur, though they do occur in other species. Chocolate Labradors, for instance, have the gene that dilutes black to brown, but this is very rare, if it occurs at all, in horses. Some dilution genes in horses do affect black, changing it to shades from bluish to sepia to dirty white or even nearly pure white.
The Agouti locus is known in almost all mammals. It codes for a protein that affects more than coat color, and is complex to sequence. In general, however, more red pigment is dominant to more black pigment.
The Agouti locus is given the symbol A. Agouti alleles are A with a superscript showing the particular form of the allele. Thus Aa is the symbol for recessive black, also called non-agouti. At stands for seal brown (black with tan on the inner legs, flanks and muzzle, very hard to tell from black with the mealy gene) which is also called tan-point in some mammals. AA is the symbol for bay. A+ is the so-called wild bay, where some red pigment appears on the lower legs. Note that + is always the symbol for the “wild-type” allele, that which is believed to be the predominant gene in a truly wild or ancestral population. The wild-type allele can be very rare in a domesticated population if it has been selected against.
Every horse has two alleles at each locus. If one allele is dominant to the other at the agouti locus, that is the allele that determines the color of the horse—if the extension locus allows it to. The order of dominance at the agouti locus is wild bay is dominant to all others, bay is dominant to black and tan-point but recessive to wild bay, tan-point is dominant to black but recessive to both bays, and black is recessive to the other three alleles. This means that two recessive blacks can produce only black foals, while two wild bays can produce any color they carry the genes for.
The agouti gene, by the way, was named for a South American rodent, the agouti. It was originally defined as controlling banded hair, seen in many wild animals. In fact, banded hair (black tips on red hairs) can be found on most bay horses, though you’ll need a magnifying glass and very good light to find it. Many of the darker shaded bays actually have rather deep black tips on individual hairs. In a few extreme cases, only the tips are visible in summer coat, and a bay horse may appear to be a seal brown (black with tan shading on muzzle and flanks) in summer and a definite dark bay in winter. The horse in the photograph is probably of this type.
Agouti in horses is bay. In dogs the same genetic color is sable, and in mice the standard gray color. (The yellow is very light.)
The Extension locus is given the symbol E. Again, this locus is very widespread in mammals. The wild-type allele, E+, allows the agouti alleles to be expressed. There is also a recessive allele, Ee, which suppresses the black pigment. Not completely—a horse with two Ee alleles can still have black whiskers and may have black hairs scattered throughout the coat. (In contrast, an EeEe dog has no black in the coat or whiskers, but an EeEe fox will be a typical “red fox” color.) But it will not have the black mane, tail and lower legs of a bay. In fact, an EeEe horse will be a chestnut, regardless of what may be at the Agouti locus.
E may also have two alleles dominant to the wild-type allele. These are dominant black ED and countershading, EB. (I have to say I have my doubts about countershading, though countershading on bays is well established.)
At the E locus, alleles with more black are dominant to alleles with more red. Further, the E locus can hide what is present at the A locus. An ED horse will be black regardless of what alleles are present at the A locus, and an EeEe horse will be chestnut regardless of what is present at the A locus. The word epistatic is sometimes used to define this relationship between loci—Extension is epistatic to Agouti.
Note that I am following Sponenberg, Equine Color Genetics Third Edition, plus my own observations on hair color.